phase changes

Terms from Artificial Intelligence: humans at the heart of algorithms

Page numbers are for draft copy at present; they will be replaced with correct numbers when final book is formatted. Chapter numbers are correct and will not change now.

Phase changes describe points boundaries in a time series or {[spatial data}} where there is markedly differt type of behaviour either side. or example, stock markets are often descibed as being in a Bull market (prices generally rising, albeit with day-to-day variability) or a Bear market (prices generally faling, albeit with day-to-day variability). Typically one needs a differet {[stichastic model}} for the different phases, and, ideally but more difficult, models of when phase changes occur.

Used in Chap. 14: page 222

Stock market leading up to and after Black Monday, 19th October 1987 – note esepcially the phase change bdetween Bull and Bear markets (Source: Wikimedia).